While understanding the trade off between tumor control probability and normal tissue complication probability lies at the heart of radiotherapy treatment planning, the reporting of the dependence of complications on dose distributions is currently inadequate. This proposal will develop new methods for reporting and combining results from individual treatment protocols that are simple to use and as unbiased and comprehensive a description of the patient treatment as possible. These methods are termed atlases of radiotherapy outcomes statistics and dose distribution variable. The atlas method aims to comprehensively display dose distribution, medical cofactors, and outcomes statistics from radiation treatments in a manner that allows concise publication, combination with data from other institutions, and thorough analysis of dependence of the outcome endpoint on the variables displayed. In consequence, publication of atlases facilitates in depth meta-analysis of dependence of published outcome data on atlas variables; atlases allow for useful publication of treatment series with few to no complications; atlases facilitate the quantitative assessment of the potential risks and benefits of future treatments. To demonstrate the use of these tools, biophysical models and corresponding atlases will be used to analyze and display dose distribution dependence of outcome from our hypo- and single fraction radiotherapy of paraspinal, lung and liver tumors, and conventionally fractionated radiotherapy of prostate and lung. Severe pneumonitis and dose-volume histogram data from non-small-cell lung cancer patients from three institutions will be used to validate atlases as a method of data pooling and overcoming limitations of poor statistics in the biophysical modeling of complication probabilities. Atlases will be developed to that summarize dose distribution and complication data in the presence of treatment uncertainties. These atlases will be used in biophysical modeling of outcome data and the results compared with those from an exact method of including treatment uncertainties. Results will be published in atlas form to enable subsequent investigators to combine our data with their own, a crucial property in the context of the low numbers of complications usually found in individual treatment protocols. NIH and RTOG standards compliant public domain software will be created for atlas creation, maintenance and display. The adoption of atlases as a standard for reporting outcomes of radiotherapy treatment protocols has the potential to create a synergistic increase in the understanding of the dependence of outcome on dose distributions. This should permit the safe delivery of higher biologically effective doses to tumors, resulting in improved local control.